- Volume 78, Issue 7, 1997

Hepatitis C virus (HCV) exists in vivo as a highly variable mixture of closely related genomes (quasispecies), but the pathogenetic significance of such heterogeneity is still largely unknown. To investigate this issue, we compared the composition of HCV quasispecies found in the liver, peripheral blood mononuclear cells (PBMC) and plasma of ten patients by single-strand conformation polymorphism analysis of the E2/NS1 region and sequencing of the variants detected. We found considerable quasispecies differences between the liver and PBMC in all the patients, involving variant numbers, relative quantities and relative electrophoretic mobilities, but no apparent tissue-specific trend. Genome variants present in the liver and/or PBMC were not detected in the corresponding plasma samples, while certain HCV variants were present only in plasma. No dominant amino acids or amino acid pattern characteristic of variants present solely in the PBMC were detected in the E2/NS1 region sequenced.

Hepatitis C virus (HCV) is a major cause of chronic hepatitis worldwide, which finally leads to development of hepatocellular carcinoma. Chronic hepatitis C is characterized by several histological features in the liver which discriminate it from other forms of hepatitis: bile duct damage, lymphoid follicles and steatosis (fatty change). Little is known, however, about the role of HCV or its viral proteins in the pathogenesis of hepatitis. Recently, the core protein of HCV has been suggested to have a transcriptional regulatory function, and thereby to be involved in inducing phenotypic changes in hepatocytes. To clarify whether or not the HCV core protein has an effect on pathological phenotypes in the liver, two independent transgenic mouse lines carrying the HCV core gene were established. These mice developed progressive hepatic steatosis, indicating that the HCV core protein plays a direct role in the development of hepatic steatosis, which characterizes hepatitis C. This transgenic mouse system would be a good animal model for the study of pathogenesis in human HCV infection.

We have investigated the ability of different subgenomic fragments to reproduce the phylogenetic relationships observed between six complete genome sequences of GBV-C/hepatitis G virus (HGV). While similar relationships were observed following analysis of part of the 5′ non-coding region (5′NCR), for the coding region they were not accurately reproduced for some large fragments or for the majority of fragments of 300 or 600 nucleotides. Analysis of 5′NCR sequences from a large number of isolates, including newly obtained sequences from Pakistan, Zaire and Scotland, pro duced separate groupings of Asian, African and European/North American variants. These groupings are associated with specific polymorphisms in the 5′NCR, many of which were covariant and consistent with a proposed secondary structure for this region. The relatively low level of amino acid sequence variation observed between these geographically and phylogenetically defined groups of variants suggests that they are unlikely to display significant biological differences.

A genetic algorithm-based RNA secondary structure prediction was combined with comparative sequence analysis to construct models of folding for the distal 380 nucleotides of the 3′-untranslated region (3-UTR) of yellow fever virus (YFV). A number of structural elements that are thermodynamically stable, conserved in shape, and confirmed by compensatory mutations were revealed. At the same time structural polymorphisms were observed among strains of YFV. These polymorphisms showed an association with virulence: all wild and pathogenic strains were likely to be folded in a signifi-cantly different way from vaccine strains with reduced virulence. Structural divergence was also found among vaccine strains, with 17DD, the most virulent in the mouse model, exhibiting an intermediate pattern of folding, combining structural features of both wild and vaccine strains. The observation of a strong association between secondary structure of the 3′-UTR and virulence of YFV may help elucidate the molecular mechanisms of virus attenuation and lead to new strategies of vaccine development directed towards rational modification of secondary structure of the 3′-UTR.

The nonstructural region of the genome of the avirulent A7 strain of Semliki Forest virus (SFV) has been sequenced, so that the complete nucleotide sequence is available. Compared to the virulent SFV4 strain (produced from the infectious clone pSP6-SFV4), A7 contains 226 nucleotide changes in the translated region, which result in 47 amino acid changes. The 5′ nontranslated region has two nucleotide changes, and the 3′ nontranslated region is longer in A7 than SFV4, and contains divergent and repeated sequences. Chimeras containing SFV4 and A7 sequences and an infectious clone of A7, pSP6-CA7, were constructed. The virulence of these was tested by intraperitoneal and intranasal infection of adult BALB/c mice. It was shown that determination of the avirulent phenotype of A7 was polygenic, and required the additive effect of sequences from both the structural and nonstructural regions of the SFV genome.

We have investigated the induction of cell death in cultured cells by the virulent SFV4 and avirulent A7 strains of Semliki Forest virus (SFV). In BHK cells, death occurred by a typical apoptotic mechanism, as did the death of oligodendrocytes in glial cell cultures. For cerebellar neuron cultures, virus- induced death was due to necrosis. Although the SFV4 and A7 strains did not differ in the mechanism of induction of cell death, the virulent SFV4 strain did multiply to a higher titre in cultured neurons than the avirulent A7 strain. This is consistent with previous animal studies which indicate that the virulence of SFV strains is controlled by rapidity of multiplication in the CNS, leading to a lethal threshold of damage, rather than differential cell tropism or cell death mechanisms. The immune- mediated demyelination induced by avirulent strains may be triggered by apoptosis of oligodendrocytes, the consequences of which are obscured by death for virulent strains.

Using a eukaryotic expression system in combination with a monoclonal antibody (MAb) capable of blocking hepatitis A virus (HAV) adsorption, a cDNA clone was selected from a library of S.la/Ve-1 cells, a cell line that is highly susceptible to the virus. Sequence analysis of the cDNA revealed a single open reading frame that encoded a protein consisting of 460 amino acids. The deduced primary structure of the protein included a signal sequence, a transmembrane domain, four sites for ^-linked glycosylation, cysteine residues attributable to an immunoglobulin domain and threonine clusters characteristic of mucin-like protein. By employing a vaccinia virus expression vector, the cDNA was expressed in HeLa cells where it induced marked HAV attachment which was specifically blocked by the MAb. The cDNA obtained was thus assumed to encode a functional receptor for HAV.

The genome sequence of the rabies-related virus Mokola virus (genus Lyssavirus) has been completed by sequencing the L gene, which consists of 6384 nucleotides encoding a 2127 amino acid polymerase. Alignment of the Mokola virus L protein with other polymerases from the virus order Mononegavirales defined three domains: a divergent NH2-terminal domain, a highly conserved central domain carrying most of the functional motifs and a COOH- terminal domain with alternating conserved and divergent regions. A statistical study outlined the stringency of conservation of glycine, acidic (D, E) and basic (K, R, H) amino acids in polymerases, particularly as key residues of the conserved motifs.

We have investigated the class I-restricted CTL response specific for measles virus haemagglutinin (HA) in the spleens of mice immunized by various mucosal routes with a DNA plasmid carrying the HA gene (pV1j-HA). A single immunization with recombinant DNA injected in the buccal mucosa induced an HA-specific CTL response. Similarly, nasal immunization with the DNA vaccine induced primary CTLs against measles virus HA. Booster immunization did not enhance the CTL activity. Oral or intrajejunal immunization with the plasmid induced a CTL response of lower magnitude. However, this could be potentiated by co-administration of the mucosal adjuvant cholera toxin or cationic lipids (DOTAP). These data show that a CTL response can be generated by mucosal vaccination using DNA vaccines.

Sequence analysis was performed on 20 measles virus (MV) isolates from South Africa, five of which were obtained between 1986 and 1989 and 15 isolates collected during the 1994/95 measles season. A 590 bp fragment of the carboxyl terminus of the nucleocapsid (N) was amplified by PCR and subjected to sequence and phylogenetic analysis. Comparison of the South African MV strains with those previously described revealed that at least two distinct groups of wild-type (wt) MV exist, one of which has been circulating since 1986. The major genotype (I) was represented by the more recent isolates which showed three characteristic amino acid substitutions. Furthermore, three vaccine-like viruses with sequences very similar to the Edmonston wt strain were identified. Phylogenetic analysis of 100 MV strains allowed the assignment of new definitions for MV genotypes and subgroups. Employing these definitions, the majority of South African isolates analysed here formed a new genotype.

The RNA-binding activity of influenza A virus M1 protein was studied by cross-linking the protein to viral RNA followed by sequence analysis of the oligoribonucleotide bound to the protein as well as sequence analysis of the M1 peptide bound to the RNA. M1 was found to bind to RNA without any RNA sequence specificity, as verified in a series of filterbinding experiments using a large variety of nucleic acids including DNA. The peptide sequence that bound to the RNA was the RKLKR nuclear localization signal of M1. Site-specific mutagenesis of recombinant M1 showed that most of the basic residues in that region had to be mutated in order to inhibit RNA-binding. We also constructed an M1 mutant that no longer bound to RNA but which was still able to inhibit the in vitro transcription activity of isolated viral ribonucleoprotein, albeit to a lower extent. Mutation of the zinc-binding sequence had no effect on RNA-binding or transcription-inhibition activity.

Trehalose 6,6′-dimycolate (TDM), an immunomodu- lator, potentiates non-specific resistance in mice to influenza virus infection. When mice were injected intravenously with TDM, the striking proliferation of a minority of T-lymphocytes bearing gamma/delta T-cell receptors (γδ T-cells) that accumulated in granulomatous lungs was thought to be associated with the maintenance of acquired resistance to lethal influenza virus infection. To clarify the cellular basis of the defence against influenza virus, mice were depleted of γδ T-cells, alpha/beta (ap) T-cells, or natural killer (NK) cells by in vivo administration of corresponding antibodies prior to influenza virus infection. The depletion of γδ T-cells significantly abrogated the augmented resistance of TDM- treated mice to infection, as did depletion of either αβ T-cells or NK cells. To gain insight into the functional ability of γδ T-cells, we evaluated the cytotoxic activity ofthisT-cell subset againsta panel of target cell lines that were stably transfected with the influenza virus haemagglutinin (HA) gene from A/PR/8/34(H1N1) and A/Aichi/2/68(H3N2) strains. The γδ T-cells from TDM-treated mice showed profound cytotoxicity against the target cells expressing HA of either the H1 or H3 subtype, in a non-major histocompatibility complex-restricted manner. Taken together, these results indicate that γδ T-cells play a non-specific role, in conjunction with αβ T-cells and NK cells, in protecting mice against influenza virus infection, and that the recognition and destruction of HA-expressing target cells by the activated γδ T-cells is one of the steps involved in this anti-influenza virus immuno- surveillance.

The Junin virus strain Candid #1 was developed as a live attenuated vaccine for Argentine haemorrhagic fever. In this paper we report the nucleotide sequences of S RNA of Candid #1 and its more virulent ancestors XJ#44 and XJ (prototype). Their relationship to Junin virus wild-type MC2 strain and other closely and distantly related arenaviruses was also examined. Comparisons of the nucleotide and amino acid sequences of N and GPC genes from Candid #1 and its progenitor strains revealed some changes that are unique to the vaccine strain. These changes could be provisionally associated with the attenuated phenotype.

An established mouse model was used to evaluate the effectiveness of the major outer core protein of African horsesickness virus (AHSV), VP7, as a subunit vaccine. Adult female BALB/c mice were immunized with VP7 crystals purified from BHK cells infected with AHSV serotype 9 (AHSV-9), using three inoculations in Freund’s adjuvant. Eighty to one hundred per cent of the immunized mice were protected against a heterologous challenge with a known lethal dose of AHSV-7. The protected immunized mice did not develop any clinical signs characteristic of virulent AHSV infection in this model during the study. In contrast, 80–100% mortality was observed in the non-immunized mice that received the same challenge virus. Subsequent studies indicated that a single inoculation of 1·5 µg purified AHSV VP7 in Freund’s complete adjuvant was sufficient to protect at least 90% of mice from AHSV-7 challenge. If the antigen was presented in the absence of Freund’s complete adjuvant, 70% of the mice were still protected by one inoculation of VP7 crystals. Titres of circulating antibody against AHSV VP7, determined by competitive ELISA, did not appear to correlate with protection and passive antibody transfer from immunized BALB/c mice failed to protect syngeneic recipients from AHSV-7 challenge. Therefore, the observed protection is unlikely to be due to an antibody-mediated immune response. The number of viraemic mice and the duration of viraemia post-challenge was significantly reduced in vaccinated mice compared to non- vaccinated controls. However, the levels of viraemia were similar.

The importance of VP7 of bluetongue virus (BTV) in the binding of BTV to membrane proteins of the BTV vector Culicoides variipennis was investigated. Core BTV particles, prepared from whole viruses, lacked outer proteins VP2 and VP5 and had VP7 exposed. More core particles and whole viruses bound to membrane preparations of adults of C. variipennis and KC cells, which were cultured from this vector insect, than to membrane preparations of Manduca sexta larvae. More core particles than whole viruses bound to membrane preparations of adults of C. variipennis and KC cells. Polyclonal anti-idiotypic antibodies (anti-Id), which were made against an antigen-combining region of an anti-BTV-10 VP7 antibody and functionally mimicked VP7, bound more to the membrane preparations of adults of C. variipennis and KC cells, and less to cytosol preparations. In Western overlay analysis, the Culicoides plasma membrane preparation reduced binding of an anti-VP7 monoclonal antibody to VP7. Whole and core BTV particles and the anti-Id bound to a membrane protein with a molecular mass of 23 kDa that was present predominantly in membrane preparations of adults of C. variipennis and KC cells. This protein was present in much lower concentrations in membrane preparations of C6/36 and DM2 insect cells.

Much attention is presently focused on the quality of the immune response produced by helper T or regulatory cells because of its implications for vaccine development and immunomodulation. Glycoprotein B (gB) of herpes simplex virus (HSV) has been shown to induce a protective T cell response. To further characterize the nature of the T cell response, oligopeptides were expressed from the open reading frame of gB from HSV-2 (gB-2) as fusion proteins with β-galactosidase (GZ) in E. coli. After immunopurification using an anti-GZ affinity column, oligopeptides p59 and p65, spanning amino acid residues 339-394 and 424-484 of gB- 2 respectively, were examined for immunogenic response by delayed type hypersensitivity (DTH) in vivo and for antigenic response by T cell proliferation in vitro. p59 but not p65 was able to prime for both DTH and proliferative T cell response to whole HSV-2 and protect against challenge infection. However, when mice were pretreated with cyclophosphamide, p65 primed for a strong DTH response to a level similar to that induced by p59 in mice either pretreated or not treated with cyclophosphamide. This suggests that p65 contains epitopes capable of inducing both DTH and immunosuppression. Thus, when mice were primed with p65 before immunizing with HSV-2, their in vitro HSV- specific proliferative response was suppressed. Therefore, p59 is a good immunogen able to induce significant, though incomplete, protection. It could be considered for inclusion in a cocktail of subunit vaccines against HSV-2 whereas p65 or parts thereof should be excluded for this purpose.

The scaffolding protein and associated protease of the human herpesvirus varicella-zoster virus (VZV), encoded by genes 33·55 and 33 respectively, were synthesized in insect cells using a baculovirus expression system. The expressed 33·55 product formed numerous long, flexible, hollow rods, and in this respect differed from the herpes simplex virus type 1 (HSV-1) homologue which forms large aggregates consisting mainly of fibrous material interspersed with scaffold-like particles. Removal of 27 amino acids from the carboxy terminus of the VZV scaffolding protein by the gene 33 protease or expression of the cleaved product did not result in any discernible change in the morphology of the scaffolding protein. Again, this was in marked contrast to the situation in HSV-1 where removal of the 25 carboxy-terminal amino acids from the scaffolding protein by the associated protease or expression of VP22a results in the formation of large numbers of scaffold-like particles. Despite these differences, when cells were multiply infected with baculoviruses expressing the HSV-1 capsid shell proteins and the VZV scaffolding protein complete capsids were observed, suggesting that the VZV protein could act as a scaffold for the assembly of the HSV-1 capsid shell. The efficiency of capsid assembly was increased substantially by exchanging the 23 carboxy-terminal amino acids of the VZV scaffolding protein for the corresponding 22 carboxy-terminal amino acids of the HSV-1 homologue, supporting previous work which showed that this region was critical for the formation of intact capsids.

To define structural elements involved in translocation of human cytomegalovirus (HCMV) glycoprotein B (gB) to the inner nuclear membrane (INM) compartment, mutagenized gB derivatives with deletions of the potential membrane anchor domains or of portions of the cytoplasmic tail were stably expressed in human astrocytoma cells. Subcellular localization examined by immunofluorescence and cell fractionation suggested that all gB derivatives reached the INM; however, reduced amounts were found after deletion of the extreme carboxy terminus [amino acids 856–906; gB(Del3)]. Pulse-chase analysis revealed accumulation in nuclear fractions of all gB derivatives during the chase, except for gB(Del3), which exhibited impaired nuclear retention. A carboxy-terminal nucleoplasmin-like signal localized within the respective deletion may thus be involved in nuclear transport and retention of HCMV gB. Immunoprecipitation after 32P-radiolabelling of the gB transfectants verified that the gB molecule is phosphorylated at a carboxy-terminal consensus motif for casein kinase II.

We have developed a number of replication defective adenoviral (Ad) vectors which express transgenes under the control of the human cytomegalovirus (HCMV) immediate early (IE) gene promoter. The orientation of the expression cassette replacing E1 in the vector backbone had a significant effect on the level of transgene expression, with vectors containing expression cassettes directed towards the right end of the Ad genome expressing 7-fold higher levels of β-galactosidase (β-gal) than those with inserts in the opposite orientation. Murine cells infected with any of several different Ad vectors in which transgene expression was under the control of the HCMV IE promoter produced 10–100-fold less transgene product (such as β-gal) than similarly infected human cells. Replacing the HCMV IE promoter with the murine CMV (MCMV) IE promoter resulted in an increase in the levels of β-gal produced in murine and rat cells by approximately 5–30-fold compared to levels obtained with the HCMV IE promoter, and levels produced in human cells were the same or greater using the MCMV IE promoter compared to the HCMV IE promoter. Similar results were obtained using a luciferase reporter gene. The MCMV IE promoter, therefore, was able to drive high levels of expression without the pronounced species preferences observed for the HCMV IE promoter. The MCMV IE promoter also directed high levels of expression in vivo, suggesting that Ad vectors carrying the MCMV IE promoter may be more effective than those with the HCMV IE promoter for transgene expression in animal models.